In this application, we propose to study behavioral profiles and genetic bases of severe reading impairment. Specifically, we will establish a unique sample of 500 severely affected elementary schools children ascertained through the PMRN database and will then recruit at least two first-degree relatives of the probands for a sample of ~1,500 individuals. We will administer a comprehensive behavioral assessment to all elementary school children and collect DNA samples from all participants. The behavioral phenotypes will be comprehensively studied, both cross-sectionally and longitudinally. In addition, we propose to conduct a relatively narrowly targeted, but in-depth, molecular-genetic study of Specific Reading Disability (SRD). Specifically, we aim to evaluate, in the newly collected PMRN database samples, the association between specific candidate genes and SRD. We propose to begin this work with a gene currently under examination, KIAA0319, and anticipate that during the life of this project there will be other candidate genes put forward through the efforts of different research groups around the world. In investigating these associations, we propose to crystallize a data-analytic approach that permits simultaneous analyses of multivariate phenotypes and multiple QTLs both cross-sectionally and longitudinally. In addition, although relatively small in magnitude, this study will offer a unique prospective on the contribution of each of the candidate genes by considering identified risk and/or protective alleles and risk and/or protective haplotypes in global genetic variation and evolutionary contexts. Specifically, the candidate genes will be investigated for ancestral alleles and haplotypes and global variation of allele and haplotype frequencies in samples from 38 world populations and a number of primate species. This in-depth analysis will permit evaluation of the frequency and structure of the candidate genes'haplotype around the world and, therefore, will increase the generalizability of results indicating the presence of association. In summary, we propose to combine behavior analyses and statistical, molecular, and population genetics in an attempt to understand associations between specific candidate genes and multiple facets of SRD in a unique, phenotypically informative, large sample of trios of first-degree relatives ascertained through probands whose severity of reading impairment puts their performance below the 3rd percentile on indicators of single-word processing.